A sorting scheduling model and its optimization solution method for a suspension line automatic sorting system
By modeling the suspended line automatic sorting system and applying genetic algorithm optimization, the problem of low sorting efficiency is solved, a more efficient sorting process is achieved, and cost is reduced.
Patent Information
- Application Number
- CN202210244965.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2022-03-13
- Publication Date
- 2024-11-19
- Estimated Expiration
- 2042-03-13
AI Technical Summary
In actual application, the automatic sorting system of hanging line has idle waiting for sorting and packaging ports, resulting in low sorting efficiency and affecting delivery timeliness.
By abstracting the hanging line automatic sorting system into a mathematical model and using a genetic algorithm based on elite retention strategy for optimization and solution, the allocation and packaging order of orders and sorting ports are optimized to improve sorting efficiency.
It effectively improves the sorting efficiency of the suspended line automatic sorting system, reduces the time for sorting all orders, and reduces the overall cost of the system.
Smart Images

Figure CN114595983B_ABST
Abstract
Description
Technical Field
[0001] The invention belongs to the field of modern logistics warehouse management, and in particular relates to a sorting scheduling model of a hanging line automatic sorting system and an optimization solution method thereof. Technical Background
[0002] With the development of the market, customers have higher and higher personalized requirements for delivery, which are mainly reflected in: the timeliness of goods delivery is constantly increasing, and next-day or even same-day delivery has become the goal of many e-commerce and postal service providers; there are more and more orders of multiple varieties and small batches. Logistics distribution centers are derived from the continuous changes and evolution of warehouses to adapt to this new economic environment and take on the delivery tasks of enterprises.
[0003] In the past, the sorting task of goods was mainly completed manually in many distribution centers, but manual sorting also has major drawbacks, such as high labor intensity of workers, low sorting efficiency, high sorting error rate, and increasing labor costs in recent years. With the changes in the market environment, manual sorting has long been unable to meet the needs. In order to efficiently handle such a large amount of logistics in a short time, the application of automated logistics equipment has played a more critical role. The automatic sorting system has the characteristics of high sorting efficiency, high precision, and small error. At the same time, there are often some heavier goods in the distribution center. If they are handled manually, they are easy to cause harm to the human body for a long time. Automated equipment can just solve this problem. At present, my country is slowly entering an aging society, and automated equipment is needed to replace manual processing. Today, many supply chains regard automatic sorting systems as an important part of the entire process.
[0004] In recent years, with the development of various industries in my country, especially the e-commerce industry, the logistics industry has also received full attention and has developed rapidly. The huge logistics business volume has put forward extremely high requirements on the sorting capabilities of logistics distribution centers and warehouses of various express companies and e-commerce platforms. Since the advent of the automatic sorting system in the 1920s, the automatic sorting system has become an indispensable part of the logistics industry and will also have a bright development prospect.
[0005] Modularization and flexible layout are also a major development trend of automatic sorting systems. Traditional automatic sorting systems are large in size and occupy a large area. In recent years, the price of commercial land has soared, which has increased the overall logistics costs. At the same time, it is difficult to maintain and the cost is high. Therefore, modularization and flexible layout have gradually become an important focus of many equipment R&D manufacturers. As a new generation of highly scalable and flexible automatic sorting systems, the hanging line sorting system has emerged and has been favored by many companies, especially clothing companies and small commodity e-commerce companies in recent years.
[0006] However, there are still some problems in the actual application of the suspension line system, such as idle waiting at the sorting and packaging port, which affects efficiency, resulting in low sorting efficiency in the actual application process. For logistics centers with large shipments, such a phenomenon should be avoided as much as possible, because it will greatly affect the timeliness of its shipments. Therefore, optimizing the sorting and scheduling of the system has important technical and economic value. Summary of the invention
[0007] In view of the above technical problems existing in the prior art, the purpose of the present invention is to provide a sorting scheduling model and an optimization solution method for a suspension line automatic sorting system. The present invention mainly abstracts the system into a mathematical model and improves the overall sorting efficiency of the system through an optimization algorithm.
[0008] A sorting scheduling model for a hanging line automatic sorting system, characterized in that the modeling process is as follows:
[0009] 1) Set the structure of the hanging line automatic sorting system
[0010] The suspension line system includes a plurality of loop lines, a conveying line and a main sorting line. The system structure is as follows: Figure 1 As shown, it consists of multiple "rings".
[0011] The group of circular lines on the left serve as temporary storage. Each circular line on the left stores a product. They are connected to the storage area and can be replenished from the storage area.
[0012] Among the multiple circular lines that constitute a group, each circular line has a number of hanging units for storing products suspended along its moving circular line, the same product is stored in the hanging units of each circular line, and different circular lines store different products. Each circular line is connected to a storage area for replenishing the products from the storage area; during the operation of the system, the products stored in the hanging units on each circular line are constantly moving around the ring, which serves as a temporary storage of the products; during the operation of the system, the products on each "ring" are constantly moving around the ring.
[0013] The large ring on the right is the main sorting line, which has multiple sorting and packaging ports at different positions along the line edge in the circular direction, for sorting and packaging the products output from each circular line of the suspension line system;
[0014] 2) Set the shipping rules for each ring line of the suspension line system
[0015] The hanging units storing the products are output from each circular line of the hanging line system, and the output hanging units are sent to the main sorting line along the conveyor line. Then the main sorting line sends the products in the hanging units to the sorting and packaging port at the side of the line for packaging the order. Each sorting and packaging port can only package one order at a time. Therefore, by setting the shipping rules of each circular line, the order and quantity of the products entering the main sorting line can be controlled. Among them, when packaging an order, each circular line outputs the products in reverse order. According to the distance from the main sorting line from far to near, different products are continuously output to the main sorting line in sequence along the same conveyor line. In this way, by controlling the time point and quantity of product output from each circular line, all products contained in the order can be connected together adjacent to each other with the minimum spacing, so that the head-to-tail distance of the entire order (that is, the distance from the first product to the last product of the order during the transportation process) is the shortest.
[0016] That is, when receiving the system instruction, "turn out" the required product from the left circular line, press Figure 1 The products are sent to the main sorting line on the right in the direction of the middle arrow, and then the main sorting line sends the products to the sorting and packaging port on the side of the line for packaging. Each sorting and packaging port can only pack one order at a time. Therefore, by controlling the circular line on the left, the order and quantity of the products entering the main sorting line can be controlled.
[0017] The left circular line outputs products in reverse order. Figure 2 As shown in Figures 3 and 4, an order contains three products, A, B, and C. The leftmost circular line carrying product A first continuously outputs the product A required for the order and transports it to the right along the conveyor line below; when the last product A is about to pass through the circular line carrying product B, product B is continuously output immediately; then when the last product B is about to pass through the circular line carrying product C, product C is continuously output immediately. In this way, by controlling the time point and quantity of each circular line outputting products, all products contained in the order can be connected together adjacent to each other with the minimum spacing, so that the length of the entire order is the shortest.
[0018] 3) Define the symbols of parameters and variables used in the specific mathematical model:
[0019] J: order set: J = {1, ..., J}, the number of orders is n;
[0020] I: sorting and packing port set: I = {1, ..., I}, the number of sorting and packing ports is m;
[0021] S ij : indicates the time when order j starts to be packed at sorting and packing port i;
[0022] T ij: represents the time required for order j to be packed at sorting and packing port i
[0023] C max : Indicates the maximum completion time after the sorting and packing port has processed all orders;
[0024] k j : represents the quantity of products in order j;
[0025] p: represents the time required to sort a product in sorting and packing port i;
[0026] q: represents the time of changing boxes after the last order is processed in sorting and packing port i;
[0027] v: represents the moving speed of the suspension unit;
[0028] d: The minimum distance between adjacent suspension units when they move in the suspension line system;
[0029] x ijt Indicates that if order j is the tth order to be packed at sorting and packing port i, it is 1, otherwise it is 0;
[0030] 4) Determine the objective function of the sorting scheduling model of the hanging line sorting system:
[0031] In the present invention, when the system receives an order, it will automatically generate an order sorting task. The left circular line will ship according to the order sequence generated by the algorithm and the rules in step 2), and the products of the same order will enter the right main sorting line for sorting and packaging. For the entire automatic sorting system, it is necessary to improve the sorting efficiency by optimizing the allocation of orders and sorting and packaging ports, and determining the packaging order of orders in each sorting and packaging port. Ultimately, it is necessary to minimize the time to sort all orders, that is, minimize the maximum sorting time:
[0032] The objective function expression is as follows:
[0033] Min C max (1)
[0034] 5) Constraints included in the model:
[0035] Obviously, the maximum sorting completion time must be greater than or equal to the sorting completion time of any order, that is:
[0036]
[0037] Among them, the sorting and packaging time of each order is related to the number of products contained in the order, the box changing time at the sorting and packaging port, and the moving speed of the hanging unit (the speed of the hanging unit on the loop line, conveyor line and main sorting line is the same), that is:
[0038]
[0039] In the model, the time when any order starts to be packaged must be greater than or equal to 0, so:
[0040] S ij ≥0 (4)
[0041] Each sorting and packing port can only pack one order at a time, and in each sorting and packing port, the start time of packing the next order must be greater than or equal to the end time of packing the previous order.
[0042]
[0043] Each order needs to be sorted and packed, and can only be packed once:
[0044]
[0045] Finally, the decision variable x ijt Can only be 0 or 1.
[0046] x ijt ∈{0,1} (8)
[0047] The present invention provides an optimization solution method for a sorting scheduling model of a suspension line automatic sorting system, which is characterized by using a genetic algorithm based on an elite retention strategy to solve the model, and includes the following process:
[0048] The present invention establishes a 0-1 integer programming model for the sorting scheduling problem of the hanging line automatic sorting system, and uses a genetic algorithm to effectively solve it in a short time. The genetic algorithm is an intelligent optimization algorithm that draws on the genetic idea of survival of the fittest in biology, and its optimization and solution process draws on the process of chromosome evolution in biology.
[0049] Genetic algorithms have been widely used in the study of scheduling problems, and the results are good. The steps of genetic algorithms to solve problems can generally be divided into the following parts: encoding, initializing the population, calculating the fitness function, selection, crossover, and mutation. The basic ideas of genetic algorithms for solving problems are as follows:
[0050] a. Determine the encoding method of the problem solution;
[0051] b. Initialize the population and obtain a set of initial feasible solutions;
[0052] c. Calculate the fitness function value of each individual in the current population;
[0053] d. Apply selection, crossover, and mutation to obtain a new generation of population, and calculate the fitness value for each individual in the new generation of population;
[0054] e. Continue to iterate until the specified number of iterations is met and the fitness value tends to be stable.
[0055] Next, a genetic algorithm is used to optimize and solve the problem raised by the present invention.
[0056] The genetic algorithm process designed by the present invention is as follows Figure 5 As shown, the specific steps include:
[0057] 1) Encoding
[0058] Coding is the most basic step in the problem-solving process of genetic algorithms. Coding is the mapping of phenotypes to individual genotypes. The basis for genetic algorithms to solve a variety of optimization problems is to select and design different coding methods according to different problems. Therefore, how to design a method that is applicable to practical problems and can fully express the solution to the problem, so that all solutions generated in the early stage of problem solving and the iteration process of the population are feasible solutions to the problem, will be the key to affecting the subsequent series of selection, crossover, and mutation operations. In addition, the coding method can affect the ability of genetic algorithms to solve and optimize problems. Good coding design can significantly improve the performance of genetic algorithms.
[0059] In the solution space of the problem studied by the present invention, each chromosome represents a feasible scheduling solution. In each scheduling solution, what needs to be solved are: ① the allocation of orders to sorting and packaging ports; ② the order of orders packaged at each sorting and packaging port. Therefore, the chromosome encoding needs to reflect these two aspects.
[0060] The present invention adopts a real number coding method. The length of the chromosome indicates the number of orders to be scheduled. For example, if the length of the chromosome is 100, it means that there are 100 orders to be scheduled. The chromosome gene position number corresponds to the order number. At the same time, the number on each gene position on the chromosome is the number of the sorting and packaging port, indicating that the order corresponding to the gene position is assigned to the sorting and packaging port with the number, and the number of times the number on a gene position on the chromosome appears indicates the number of times the order corresponding to the gene position is packed at the sorting and packaging port with the number.
[0061] For example, Figure 6 Chromosome shown: The length of the chromosome is 10, indicating that there are 10 orders. Order 1 and order 8 are packaged at sorting and packaging port 1, and order 1 is packaged before order 8. Similarly, order 2 and order 5 are packaged at sorting and packaging port 3 in sequence; order 3 is packaged at sorting and packaging port 4; order 4, order 6 and order 10 are packaged at sorting and packaging port 5 in sequence; order 7 and order 9 are packaged at sorting and packaging port 2 in sequence.
[0062] 2) Initialize the population
[0063] In the biological world, population is the basic unit of biological evolution. The same is true in genetic algorithms, which require continuous iteration of the population to find the optimal solution. The population size is designed to be N, and the population consists of a set of chromosomes. Since the encoding method of the chromosomes has been determined, a set of initial feasible solutions, namely the population, is obtained by randomly generating the number of the sorting and packaging port at each gene position.
[0064] At the same time, it is worth noting that the size of the population will directly affect the efficiency and performance of the genetic algorithm in the optimization process. Specifically, if the population size is too small, it will not be able to provide enough chromosome individuals, making the genetic algorithm prone to "premature maturity" and the resulting solution poor, making it difficult to obtain the optimal solution or the optimal approximate solution; conversely, if the population size is too small, although local convergence can be avoided, it will inevitably greatly increase the amount of calculation, thereby affecting the efficiency of the computer solution.
[0065] 3) Decoding and fitness calculation
[0066] In nature, the rule of survival of the fittest depends on the individual's ability to adapt to the environment. If an individual has a strong ability to adapt to the environment, then the individual has a higher chance of surviving and continuing to reproduce; on the contrary, if the individual's ability to adapt to the environment is weak, then the individual is likely to be eliminated by natural selection. In other words, fitness is the standard for evaluating the degree of adaptation of an individual in a population to the environment.
[0067] In a genetic algorithm, each individual has its fitness, which represents the quality of the solution corresponding to the individual. The higher the fitness function value, the better the solution corresponding to the individual, and the more likely it is to be retained for iteration in the subsequent selection step; conversely, the lower the fitness function value, the worse the solution corresponding to the individual, and the higher the probability of being eliminated in the selection step.
[0068] When using genetic algorithms to optimize and solve problems, the fitness function should match the solution goal of the specific problem. Generally speaking, optimization problems can be roughly divided into two types, one is to find the global minimum value of the objective function, and the other is to expect to find the global maximum value of the objective function. It should be noted that in the selection stage of the algorithm, the individual with a larger fitness function value has a greater chance of being selected. For example, the roulette selection method is used in this article. The larger the fitness value of the individual, the larger the corresponding roulette area and the greater the probability of being selected. Therefore, when the objective function is the global maximum value, the fitness function can be equal to the objective function; and when the objective function is the global minimum value, the objective function needs to be converted into the fitness function by methods such as finding the inverse, so that when the fitness function value is larger, the obtained objective function value is smaller.
[0069] In the present invention, during decoding, the chromosome sequence of the individual is traversed from left to right, and all the same numbers, i.e., the sorting and packing port numbers, and the corresponding order times are accumulated to obtain the working time of each sorting and packing port, and the largest one is taken as the total sorting time for this batch of orders.
[0070] In the present invention, it is expected to obtain the optimal scheduling solution so that the total time to sort all orders is minimized. Therefore, the fitness function can be taken as the inverse of the total sorting time, so that the solution with the shorter total time to sort all orders has the greater fitness value. Let f(x i ) represents the fitness function, then:
[0071]
[0072] 4) Select
[0073] The selection operation simulates the rule of "survival of the fittest" in nature. The selection operation is based on the fitness function value mentioned above. As mentioned above, the individuals with larger fitness function values have a greater probability of being selected to survive and undergo crossover mutation. In genetic algorithms, generally speaking, selection strategies can be divided into two types: the first can be called the "pure selection" strategy. The rule followed by this strategy is that the fitness of an individual is proportional to its probability of being selected. The "pure selection" methods used in selection operations are generally such as roulette selection method, tournament selection method, etc.; the other is the "optimal" strategy, which first performs "pure selection" and then inserts the current optimal individuals into the next generation population. In the practical application of genetic algorithms, these two strategies are generally used in combination.
[0074] The present invention first adopts the more commonly used roulette selection method, such as Figure 7 The basic idea of the roulette wheel selection method is to make the probability of each individual being selected proportional to its fitness.
[0075] The principle of this algorithm is actually quite simple, and its core lies in two probabilities, namely:
[0076] ① Individual selection probability:
[0077]
[0078] ②Cumulative probability:
[0079]
[0080] Where f(x i ) is individual x i The fitness function value of , N is the population size, it is easy to know that Q(x i ) has a maximum value of 1.
[0081] When making a selection, a value in the interval (0,1) is randomly generated, and then the region in which the value falls is determined. Figure 7 As shown, it is easy to see that P(x i ) is larger, the larger the area ratio of the entire roulette wheel is, and the greater the probability that a random value will fall in this area. Therefore, under this principle, the roulette algorithm realizes the “survival of the fittest” for individuals in the population.
[0082] 5) Cross
[0083] Although the selection process can bring the population closer to the optimal solution, new individuals cannot be generated simply by selecting the parent generation. Therefore, in order to improve the algorithm's search ability, expand the search space, and obtain a relatively better solution, chromosomes need to be crossovered.
[0084] Through the selection operation in the previous step, the chromosomes for the crossover operation can be selected. Through the crossover operation of chromosomes, the parent individuals can exchange some genes according to certain rules to obtain new individuals.
[0085] In the genetic algorithm of the present invention, a single-point crossover method is adopted, that is, only one crossover point is set on each chromosome, so that the chromosome can be divided into two parts, and the chromosomes cross each other to form new offspring chromosomes. Figure 8 shown.
[0086] 6) Mutation
[0087] After the crossover operation, the chromosome mutation operation needs to be performed. The mutation imitates the mutation of certain genes of chromosomes caused by accidental factors in nature.
[0088] The role of the mutation step is to maintain and protect the diversity of the established population, improve the local search ability of the algorithm, and thus avoid the premature phenomenon of the genetic algorithm. The combination of mutation operation and crossover operation can make the genetic algorithm have good local search and global search capabilities, thereby optimizing the process of finding the solution to the problem.
[0089] There are many ways to perform mutation operations in genetic algorithms. In genetic algorithms using integer coding, the more commonly used methods include swap mutation and reverse mutation. The mutation process of the present invention adopts the position swap mutation method, that is, first randomly select two different positions in a chromosome, and then swap the sorting and packaging port numbers at these two positions, such as Fig. 9 shown.
[0090] 7) Elite retention strategy
[0091] The selection, crossover, and mutation algorithms mentioned above cannot converge to the global optimum, and the solutions obtained in each generation fluctuate greatly. This is because the best individuals in the population obtained in each generation are also included in the selection, crossover, and mutation process, which causes these best individuals to disappear in the next generation and be replaced by relatively poor individuals. As the population iteration continues, this process of continuous generation and loss of the best individuals will occur repeatedly in the process of population evolution.
[0092] The elite retention strategy is a method proposed to prevent the best individuals from being lost during the iterative evolution of the population. The main idea is to retain the better individuals in each generation, the so-called "elite individuals", so that these individuals do not participate in the selection, crossover, and mutation process. When the remaining individuals complete the selection, crossover, and mutation operations to obtain a new population, the above-mentioned "elite individuals" are inserted into the new population.
[0093] In order to keep the size of each generation of population unchanged, while retaining the elite individuals, it is necessary to eliminate some of the poor individuals during the selection operation in step 4). The specific operation process is: add a "generation gap probability", that is, by selecting the ratio of the population size retained to the original population size, calculate the ratio of the ratio of the population size retained by selection to the original population size. In each generation of the population, the elite individuals with this ratio ratio will be retained and directly added to the next generation of the population, and the individuals with the worst ratio ratio will be eliminated in the next generation of the population to ensure that the overall population size remains unchanged.
[0094] For example, if the calculated ratio is 10%, then 10% of the elite individuals in each generation of the population will be retained and directly added to the next generation of the population, and 10% of the worst individuals will be eliminated from the next generation of the population to ensure that the overall population size remains unchanged.
[0095] By combining with the roulette wheel selection method, individuals with lower fitness values in the previous generation are eliminated. Finally, individuals with higher fitness values in the previous generation population are reinserted into the offspring population after crossover and mutation, so as to ensure that the better individuals in the current population can be continuously propagated to the next generation.
[0096] 8) Termination conditions
[0097] Finally, the termination condition of the algorithm iteration is designed, and the maximum number of iterations is set. The termination condition is to set the termination evolutionary generation for it.
[0098] The beneficial effects achieved by the present invention are as follows: the present invention mainly designs a sorting scheduling model and an optimization solution method for a hanging line automatic sorting system. Thanks to the e-commerce economy, my country's logistics industry has developed rapidly in recent years. In addition to e-commerce, we find that logistics has played an increasingly important role in people's production and life. The automated sorting system is an indispensable part of the development of logistics, and the importance of the automated sorting system in various industries is constantly increasing, which is specifically reflected in the fact that the automated sorting system has been adopted by more and more companies in various industries to improve their level of customer service. At the same time, new automated sorting equipment continues to emerge in the market. However, due to the wide range of logistics, this also means that the automated sorting system has to face a variety of products with different sizes, shapes, weights and other characteristics. There are many manufacturers producing automated sorting equipment, but there is still a lack of relevant standards between manufacturers. Therefore, although the automated sorting system is widely used in the industry and has played a considerable role compared with the past, the existence of the above factors makes it difficult to cover the operating modes and characteristics of most automated sorting systems with a unified model. It is worth noting that the construction and maintenance costs of the current automatic sorting system are still relatively high. Therefore, it is of great significance to improve its sorting efficiency at the lowest possible cost. Changing the system's sorting strategy is one of the important ways to improve efficiency.
[0099] Therefore, the present invention abstracts the sorting and scheduling process of the hanging line sorting system into a mathematical model, and designs a genetic algorithm to solve it according to the characteristics of the problem, which can effectively improve the sorting efficiency of the sorting system and has high technical and economic value. BRIEF DESCRIPTION OF THE DRAWINGS
[0100] Figure 1 This is a structural diagram of the hanging line sorting system;
[0101] Figure 2 This is a schematic diagram of the A product shipped along the circular line on the left;
[0102] Figure 3 This is a schematic diagram of the left circular line shipping product B;
[0103] Figure 4 This is a schematic diagram of the C product shipped along the circular line on the left;
[0104] Figure 5 is the genetic algorithm process;
[0105] Figure 6 A schematic diagram of chromosomes;
[0106] Figure 7 This is a schematic diagram of the roulette algorithm;
[0107] Figure 8 This is a schematic diagram of the chromosome single-point crossover process;
[0108] Fig. 9 This is a schematic diagram of the chromosome exchange mutation process;
[0109] Fig.10 is the demand for each product in the 50 orders in the embodiment;
[0110] Fig.11 It is a genetic algorithm iteration curve of an embodiment;
[0111] Fig.12 Gantt chart of the genetic algorithm scheduling results of the embodiment. DETAILED DESCRIPTION
[0112] In order to facilitate understanding by those skilled in the art, the present invention is further described below with reference to examples and drawings. The contents mentioned in the implementation modes are not intended to limit the present invention.
[0113] A sorting scheduling model and optimization solution method for a hanging line sorting system. The modeling process is as follows:
[0114] 1) Symbol Definition
[0115] J: order set: J = {1, ..., J}, the number of orders is n;
[0116] I: sorting and packing port set: I = {1, ..., I}, the number of sorting and packing ports is m;
[0117] S ij : indicates the time when order j starts to be packed at sorting and packing port i;
[0118] T ij : represents the time required for order j to be packed at sorting and packing port i
[0119] C max : Indicates the maximum completion time after the sorting and packing port has processed all orders;
[0120] k j : represents the quantity of products in order j;
[0121] p: represents the time required to sort a product in sorting and packing port i;
[0122] q: represents the time of changing boxes after the last order is processed in sorting and packing port i;
[0123] v: represents the moving speed of the suspension unit;
[0124] d: The minimum distance between adjacent suspension units when they move in the suspension line system;
[0125] x ijtIndicates that if order j is the tth order to be packed at sorting and packing port i, it is 1, otherwise it is 0;
[0126] 2) Model building
[0127] By analyzing the characteristics of the problem, the mathematical model is as follows:
[0128] Min C max (1)
[0129]
[0130] S ij ≥0 (4)
[0131]
[0132] x ijt ∈{0,1} (8)
[0133] Expression (1) indicates that the objective function is to minimize the maximum completion time after the sorting and packing port has processed all orders;
[0134] Expression (2) represents the expression of the maximum completion time;
[0135] Expression (3) is an expression of the time required to pack an order;
[0136] Expression (4) means that the time when all orders start to be packaged is greater than or equal to 0;
[0137] Expression (5) indicates that for the same packaging port, the start packaging time of the next order is greater than or equal to the start packaging time of the previous order plus its packaging time;
[0138] Expression (6) indicates that a sorting and packing port can only process one order at a time;
[0139] Expression (7) indicates that all orders must be sorted and packed, and can only be assigned to one sorting and packing port;
[0140] Expression (8) represents x ijt Can only be 0 or 1.
[0141] When the system receives order information, it will automatically use the designed genetic algorithm to perform calculations at the system terminal. The algorithm will automatically generate chromosomes of a certain length based on the order quantity, perform crossover mutation and elite retention, and finally output the solution results. Based on the results, corresponding cargo delivery commands will be issued to various parts of the system.
[0142] The optimization solution method of the sorting scheduling model of the hanging line automatic sorting system comprises the following steps:
[0143] 1) Encoding
[0144] First, we need to determine the characteristics of the chromosome. Each chromosome represents a solution. The chromosome encoding needs to reflect the following two aspects: ① the allocation of orders to sorting and packaging ports; ② the order of orders packaged at each sorting and packaging port;
[0145] Using real number coding, the length of the chromosome represents the number of orders to be scheduled, the chromosome gene position number corresponds to the order number, and the number on each gene position on the chromosome is the number of the sorting and packaging port, indicating that the order corresponding to the gene position is assigned to the sorting and packaging port with the number, and the number of times the number on a gene position on the chromosome appears indicates the number of times the order corresponding to the gene position is packed at the sorting and packaging port with the number;
[0146] 2) Initialize the population
[0147] The population size is designed to be N, and the population consists of a set of chromosomes. Since the encoding method of chromosomes has been determined, a set of initial feasible solutions, namely the population, is obtained by randomly generating the number of the sorting and packaging port at each gene position;
[0148] 3) Decoding and fitness calculation
[0149] When decoding, the chromosome sequence of the individual is traversed from left to right, and all the same numbers, that is, the sorting and packing port numbers, and the corresponding order times are accumulated to obtain the working time of each sorting and packing port. The largest one is taken as the total sorting time for this batch of orders;
[0150] The optimal scheduling solution is the one that takes the shortest total time to sort all orders. Therefore, the fitness function takes the inverse of the total sorting time, which can minimize the total time to sort all orders and maximize the fitness value. Let f(x i ) represents the fitness function, then:
[0151]
[0152] 4) Select
[0153] Using the roulette wheel selection method, calculate the following two probabilities:
[0154] ③Individual selection probability:
[0155]
[0156] 4 Cumulative probability:
[0157]
[0158] Where f(x i) is individual x i The fitness function value, N is the population size, Q(x i ) has a maximum value of 1;
[0159] When making a selection, a value in the interval (0,1) is randomly generated, and then the region in which the value falls is determined; P(x i ) is larger, the larger the area occupied in the entire roulette wheel is, and the greater the probability that the random value falls in this area;
[0160] 5) Cross
[0161] Through the selection operation in the previous step, the chromosomes for the crossover operation can be selected; through the crossover operation of chromosomes, the parent individuals can exchange some genes according to certain rules to obtain new individuals;
[0162] The single-point crossover method is adopted, that is, only one crossover point is set on each chromosome, so that the chromosome can be divided into two parts, and the chromosomes cross each other to form new daughter chromosomes;
[0163] 6) Mutation
[0164] After the crossover operation, the position exchange mutation method is used to perform the chromosome mutation operation, that is, firstly randomly select two different positions in a chromosome, and then exchange the sorting and packaging port numbers at these two positions;
[0165] 7) Elite retention operation
[0166] The elite retention operation is to prevent the best individuals from being lost during the iterative evolution of the population. The operation process is to retain the better individuals in each generation, that is, the elite individuals, so that these individuals do not participate in the selection, crossover, and mutation process; when the remaining individuals complete the selection, crossover, and mutation operations to obtain a new population, the elite individuals are inserted into the new population;
[0167] In order to keep the size of each generation of population unchanged, while retaining the elite individuals, it is necessary to eliminate some of the poor individuals during the selection operation in step 4). The specific operation process is: add a generation gap probability, that is, by selecting the ratio of the remaining population size to the original population size; by combining with the roulette selection method, eliminate the individuals with lower fitness values in the previous generation, and finally reinsert the individuals with higher fitness values in the previous generation population into the offspring population after crossover and mutation, so as to ensure that the better individuals in the current population can be continuously propagated to the next generation;
[0168] 8) Termination conditions
[0169] Finally, the termination condition of the algorithm iteration is designed, and the maximum number of iterations is set. The termination condition is to set the termination evolutionary generation for it.
[0170] Embodiment 1:
[0171] Example: The processing time of each product at the sorting and packaging port is 15s, the minimum distance between the hanging units is 0.10m, the distance between the sorting and packaging ports is 5m, and the conveying speed of the hanging line is 1m / s. There are 6 sorting and packaging ports, 50 orders, and the demand for each product in the order is as follows: Fig.10 In the experiment, the algorithm iteration curve is shown as Fig.11 As shown, the scheduling result Gantt chart is as follows Fig.12 shown.
[0172] The genetic algorithm parameter settings are shown in Table 1. The algorithm can converge within 200 generations.
[0173] Table 1 Genetic algorithm parameter settings
[0174]
[0175] The contents described in this specification are merely an enumeration of implementation forms of the inventive concept, and the protection scope of the present invention should not be regarded as being limited to the specific forms described in the embodiments.
Claims
1. A sorting scheduling model for a hanging line automatic sorting system, characterized in that The modeling process is as follows: 1) Set the structure of the hanging line automatic sorting system The hanging line automatic sorting system includes a plurality of loop lines constituting a group, a conveying line and a main sorting line; In the plurality of loop lines constituting a group, each loop line has a plurality of suspension units for storing products suspended along its moving loop line, the same product is stored in the suspension units of each loop line, and different products are stored in different loop lines, and each loop line is connected to a storage area for replenishing from the storage area; during the operation of the system, the products stored in the suspension units on each loop line are constantly moving around the loop, so as to temporarily store the products; The storage area is used to replenish products for each loop line of the suspension line system; The main sorting line is provided with a plurality of sorting and packing ports at different positions along the line edge in a circular direction, for sorting and packing the products outputted from each circular line of the suspension line system; 2) Set the shipping rules for each ring line of the suspension line system The hanging units storing the products are output from each circular line of the hanging line system, and the output hanging units are sent to the main sorting line along the conveyor line. Then the main sorting line sends the products in the hanging units to the sorting and packaging port at the side of the line for packaging the order. Each sorting and packaging port can only package one order at a time. Therefore, by setting the delivery rules of each circular line, the order and quantity of the products entering the main sorting line can be controlled. When packaging an order, each circular line outputs the products in reverse order. Different products are continuously and closely followed along the same conveyor line to the main sorting line according to the distance from the main sorting line from far to near. In this way, by controlling the time point and quantity of the output products of each circular line, all the products contained in the order can be connected together adjacent to each other with the minimum spacing, so that the conveying length of the overall order is the shortest. 3) Define the symbols of the parameters and variables used in the model J: order set: J = {1, ..., J}, the number of orders is n; I: sorting and packing port set: I = {1, ..., I}, the number of sorting and packing ports is m; S ij : indicates the time when order j starts to be packed at sorting and packing port i; T ij : represents the time required for order j to be packed at sorting and packing port i C max : Indicates the maximum completion time after the sorting and packing port has processed all orders; k j : represents the quantity of products in order j; p: represents the time required to sort a product in sorting and packing port i; q: represents the time of changing boxes after the last order is processed in sorting and packing port i; v: represents the moving speed of the suspension unit; d: The minimum distance between adjacent suspension units when they move in the suspension line system; x ijt Indicates that if order j is the tth order to be packed at sorting and packing port i, it is 1, otherwise it is 0; 4) Determine the objective function of the sorting scheduling model of the hanging line automatic sorting system: When the hanging line automatic sorting system receives an order, it will automatically generate an order sorting task; the order is shipped according to the rules in step 2), and the products of the same order enter the main sorting line for sorting and packaging. For the entire automatic sorting system, it is necessary to improve the sorting efficiency by optimizing the allocation of orders and sorting and packaging ports, and determining the packaging order of orders in each sorting and packaging port. Ultimately, it is necessary to minimize the time to sort all orders, that is, minimize the maximum sorting time: The objective function expression is as follows: MinC max (1) 5) The model needs to meet the following constraints: The maximum completion time must be greater than or equal to the sorting end time of any order, that is: The sorting and packing time of each order is related to the number of products contained in the order, the box changing time at the sorting and packing port, and the moving speed of the hanging unit, namely: In the model, the time when any order starts to be packed must be greater than or equal to 0, so the constraints are as follows: S ij ≥0 (4) Each sorting and packing port can only pack one order at a time, and in each sorting and packing port, the start time of the next order packing must be greater than or equal to the end time of the previous order packing. Therefore, the constraints are as follows: Each order needs to be sorted and packed, and can only be packed once: Finally, the decision variable x ijt Can only take 0 or 1; x ijt ∈{0,1} (8)。 2. The optimization solution method of the sorting scheduling model of the hanging line automatic sorting system according to claim 1 is characterized in that The following steps are involved: 1) Encoding First, we need to determine the characteristics of the chromosome. Each chromosome represents a solution. The chromosome encoding needs to reflect the following two aspects: ① the allocation of orders to sorting and packaging ports; ② the order of orders packaged at each sorting and packaging port; Using real number coding, the length of the chromosome represents the number of orders to be scheduled, the chromosome gene position number corresponds to the order number, and the number on each gene position on the chromosome is the number of the sorting and packaging port, indicating that the order corresponding to the gene position is assigned to the sorting and packaging port with the number, and the number of times the number on a gene position on the chromosome appears indicates the number of times the order corresponding to the gene position is packed at the sorting and packaging port with the number; 2) Initialize the population The population size is designed to be N, and the population consists of a set of chromosomes. Since the encoding method of chromosomes has been determined, a set of initial feasible solutions, namely the population, is obtained by randomly generating the number of the sorting and packaging port at each gene position; 3) Decoding and fitness calculation When decoding, the chromosome sequence of the individual is traversed from left to right, and all the same numbers, that is, the sorting and packing port numbers, and the corresponding order times are accumulated to obtain the working time of each sorting and packing port. The largest one is taken as the total sorting time for this batch of orders; The optimal scheduling solution is the one that takes the shortest total time to sort all orders. Therefore, the fitness function takes the inverse of the total sorting time, which can minimize the total time to sort all orders and maximize the fitness value. Let f(x i ) represents the fitness function, then: 4) Select Using the roulette wheel selection method, calculate the following two probabilities: ① Individual selection probability: ②Cumulative probability: Where f(x i ) is individual x i The fitness function value, N is the population size, Q(x i ) has a maximum value of 1; When making a selection, a value in the interval (0,1) is randomly generated, and then the region in which the value falls is determined; P(x i ) is larger, the larger the area occupied in the entire roulette wheel is, and the greater the probability that the random value falls in this area; 5) Cross Through the selection operation in the previous step, the chromosomes for the crossover operation can be selected; through the crossover operation of chromosomes, the parent individuals can exchange some genes according to certain rules to obtain new individuals; The single-point crossover method is adopted, that is, only one crossover point is set on each chromosome, so that the chromosome can be divided into two parts, and the chromosomes cross each other to form new daughter chromosomes; 6) Mutation After the crossover operation, the position exchange mutation method is used to perform the chromosome mutation operation, that is, firstly randomly select two different positions in a chromosome, and then exchange the sorting and packaging port numbers at these two positions; 7) Elite retention operation The elite retention operation is to prevent the best individuals from being lost during the iterative evolution of the population. The operation process is to retain the better individuals in each generation, that is, the elite individuals, so that these individuals do not participate in the selection, crossover, and mutation process; when the remaining individuals complete the selection, crossover, and mutation operations to obtain a new population, the elite individuals are inserted into the new population; In order to keep the size of each generation of population unchanged, while retaining the elite individuals, it is necessary to eliminate some of the poor individuals during the selection operation in step 4). The specific operation process is: add a generation gap probability, calculate the ratio of the population size retained by the selection to the original population size, so that the elite individuals with the ratio ratio in each generation of population are directly added to the next generation population, and the individuals with the worst ratio ratio are eliminated in the next generation population, so as to ensure that the overall population size remains unchanged; by combining with the roulette selection method, eliminate the individuals with lower fitness values in the previous generation, and finally reinsert the individuals with higher fitness values in the previous generation population into the offspring population after crossover and mutation, so as to ensure that the better individuals in the current population can be continuously propagated to the next generation; 8) Termination conditions Finally, the termination condition of the algorithm iteration is designed, and the maximum number of iterations is set. The termination condition is to set the termination evolutionary generation for it.
